Motion (and sound) input comes from a 3D video I made using Honey Select.
This one took quite a lot of time to make, as I wanted to make it as stable and detailed as possible when using 3D input. The facial expressions and lip syncing were almost entirely done manually.
I am putting a lot of effort into pushing the boundaries of AI videos, and I’m happy to see myself getting better over time.
Hope you can enjoy this!